Očitavanje komunalnih brojila pomoću mobilnog uređaja

master's thesis

• Title (english): Reading of Communal Meters with Mobile Devices
• Author: Zvonimir Čuljak
• Mentor: Marko Subašić (mentor)
• Committee member: Marko Subašić (predsjednik povjerenstva)
• Committee member: Sven Lončarić (član povjerenstva)
• Committee member: Mladen Vučić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-14, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom radu opisana su dva rješenja problema automatskog očitanja komunalnih brojila
pomoću mobilnog uređaja. U prvom rješenju, analiza se obavlja pomoću poslužitelja i Mask
RCNN aplikacije, dok u drugom Android aplikacija sama obavlja analizu koristeći
TensorFlow Lite i SSD MobileNets. Opisan je proces prikupljanja, označavanja i pripreme
podataka, treniranje modela i njihova evaluacija preciznosti. Razvijene su pozadinska
aplikacija za analizu, njen poslužitelj i Android korisnička aplikacija. U radu se razmatraju
rezultati, te prednosti i nedostaci ponuđenih rješenja.

Abstract (english)

This paper describes two different solutions to the problem of automatic reading of meters
using mobile devices. In the first solution, image analysis is done by a Mask RCNN backend
application connected with a server, while in the second one, Android application runs the
analysis using TensorFlow Lite and SSD MobileNets. The paper also provides insight in the
process of gathering, annotating and preparing the image dataset, training of models and
their precision evaluation. Backend image analysis app, server and an Android app also have
been developed. Furthermore, results, pros and cons of each solution is described.

• Keywords: Računalni vid; Mask RCNN; SSD; MobileNets; TensorFlow; TensorFlow Lite; Android; konvolucijske neuronske mreže; detekcija; klasifikacija; mobilna aplikacija
• Keywords (english): Computer Vision; Mask RCNN; SSD; MobileNets; TensorFlow; TensorFlow Lite; Android; Convolutional Neural Networks; Detection; Classification; Mobile Application
• Language: croatian
• URN:NBN: urn:nbn:hr:168:867634
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 09:54:01